RahulVC02/ML_Assignment-3_CNN

This Streamlit app shows a 2D CNN model trained on the ESC-10 dataset. It displays the spectrogram of the audio file and the predicted class. It also displays an animation of the saliency maps at each layer of the network. I have used the VGG16 architecture for this task. The model was trained on the ESC-10 dataset.

HTMLNOASSERTION